Paper summary: Sexuality in a community based sample of adults with autism spectrum disorder

For a while in the asexual community there’s been a lot of speculation about whether asexuality could be linked to autism.  If you’ve seen (A)Sexual, you’ll remember that one of the aces interviewed is quite adamant about there being a link between the two.

That’s why today I’m going to summarize “Sexuality in a community based sample of adults with autism spectrum disorder” by Laura Gilmour, P. Melike Schalomon, and Veronica Smith.*  This is, to my knowledge (which admittedly isn’t too extensive), the only article that specifically looks at autism and asexuality as a sexual orientation (rather than as a behavior).

I’m going to skip over some bits of the article, as they’re more pertinent to people who are interested in reading about ASD than people who are interested in reading about asexuality.  If you’d like to read the full text, you can do so here.

In the introduction, Gilmour et al. give an overview of the pre-existing literature on ASD adults and sexuality.  Most of it is not really relevant to discussing asexuality, except for these bits:

  • In a study of 89 people in group homes in North Carolina, only 34% exhibited person-oriented behavior, although masturbation was common.  Note, though, that this is person-oriented behavior, not attraction, so this tells us nothing about the actual sexual orientations of the group, only that they tended to be celibate.  In a similar study in Denmark, 74% of the group displayed person-oriented behaviors.
  • In a study conducted by Marriage et al. in 2009 on a community based sample of adults with ASD, 33% appeared to be asexual. If you look at the original paper, it says, “Anecdotally, about one third of the patients, male and female, had no interest in establishing a sexual relationship, and seemed asexual in their orientation. Some, mostly males, had tried but failed, usually because their self presentation was poor and their approach clumsy.”  Still not terribly clear, as their definition of asexuality seems, once again, to be based mostly off of behavior (although at least they did look at interest as well).
  • In past studies of individuals with ASD living in group homes, researchers have found abnormally high numbers of minority sexual orientations.  However, the sample sizes tended to be very small (20-40 people in most of the studies), and the researchers defined orientation by behavior, not by attraction, so their results should be taken with a grain of salt.

Gilmour et al. then talks about the possibility of both ASD and homosexuality being caused by prenatal androgen exposure.  If you’ve ever read Understanding Asexuality by Anthony Bogaert, you’ll also remember that prenatal androgen exposure may play a role in gender variance and (more pertinent to this summary) asexuality.  According to a 2006 article (Kraemer et al.), however, prenatal androgen exposure seems to cause homosexuality in females (I assume they mean cis women, but they don’t specify), but not males (same as before).  So Gilmour et al. posit that they will find increased levels of homosexuality in their female respondents but not in their male respondents.

It’s worth noting that Gilmour et al. are assuming that the Extreme Male Brain theory is true.  The Extreme Male Brain theory basically says that people with ASD have “extremely masculine brains,” and this theory is offered as the reason why there are so few people assigned female at birth who are diagnosed with ASD.  Gilmour et al. are assuming that the reason there will be more lesbians with ASD than gay men with ASD is because an extremely masculine brain would be more attracted to women.  (Let’s think about that logic for a second.  Um.  UM.)  The Extreme Male Brain theory has come under a great deal of criticism, not the least of which because it assumes that there is such a thing as a “gendered brain.”**  Gilmour et al. completely ignore the controversial nature of the theory, however, and base all of their hypotheses off of it.

Methods:

Data was collected from 484 participants, but 120 results were thrown out.  Of the remaining participants, 82 (55 women and 27 men) had ASD and 282 did not.

Online surveys were given to both groups that were intended to test their estimated degree of ASD, sexual experiences, and sexual orientations.

Now we get to the bit where I side-eye the heck out of their survey design:

First of all, their sample size is not all that big.  Remember, they had 27 men with ASD, and were trying to generalize from that.  They had as many men as they had participants in those 20-40 person studies that they were critiquing just a few paragraphs ago.

Second of all, their sample size was recruited from “websites which post links to research pertaining to ASD or general psychology. Examples include the Edmonton Autism Society website, North American and European sites featuring links to studies in the Social Sciences, and a personal Asperger’s Syndrome themed blog operated by one of the authors. Additional participants from the general population were recruited from a research participation site for introductory psychology students at a small Canadian University” (314-315).  Um, okay, so the autism group and the control group were recruited from ENTIRELY DIFFERENT PLACES.  And, guess what, the control group differed from the ASD group significantly not only in terms of country of origin, but also age, religious background, and race.  Wow, it’s almost like these people have no idea how to recruit a control group.  Hint to all you aspiring survey makers out there: If your control group differs substantially from your target group along axes other than the ones you’re studying, it’s not actually a control group.  And, yes, age and religiosity will have an impact on attitudes toward sex, so you should probably take all the results of this study with a block of salt.

Also, let’s talk about the survey questions.  Wow, the survey questions.  Okay, so on the survey intended to measure the sexual experience of the participants, these two questions were asked:

How often have you gone further than necking (this means doing more than kissing and hugging)?

and

Have you ever engaged in penile-vaginal intercourse (penis in vagina)?

Then they looked at the correlation between the two questions to figure out if people with ASD understood sexual language.

Um.

UM.

First of all, I need a definition of “further.”  Secondly, I have never heard “necking” defined as “hugging and kissing.”  Merriam-Webster backs me up on this.  Third of all, good to know that the only way you can go “further” than “necking” is if you have a penis and a vagina between the two of you!  This seems like an issue, given that they were trying to measure levels of non-heterosexuality.  Fourth of all, I’m pretty sure that there are a whole spectrum of things people would consider “further” than “necking” that don’t involve penises in vaginas!  Fifthly, assuming that the majority of their ASD participants was non-heterosexual and the majority of their control group was heterosexual, there is a good chance that the correlations for the two groups would be fairly different just based on personal experience, so this is a terrible way to measure knowledge of sexual language!

Gilmour et al. didn’t publish most of their survey questions in the article, but given the samples, I am extremely suspicious of their being any good AT ALL.

Results: 

Gilmour et al. found:

  • There were no significant differences in sexual experience between the two groups.
  • There were no significant differences in understanding of sexual language between the two groups
  • The ASD group scored lower on heterosexuality and higher on homosexuality, bisexuality, and asexuality than the control group.
  • Female ASD participants scored lower on heterosexuality and higher on bisexuality than male ASD participants.

Also, the results they got conflicted with Extreme Male Brain theory, which was what they were basing their hypotheses off in the first place.  Oops.

I’m just going to copy and paste the entire “Limitations” paragraph here, because it is hilarious:

One bias present in the current survey is that volunteer participants in sex surveys are likely to have a high pre-existing interest in sexuality. Our study did indicate a higher rate of asexuality among ASD participants compared to controls. However, it is possible that asexual ASD participants were still underrepresented in our study, as they would be less likely to volunteer for sex research than sexual persons with ASD. One piece of evidence supporting this contention is the fact that that the Marriage et al. (2009) study found a higher degree of asexuality in a community based sample of individuals with ASD than was found in the current sample of ASD participants. (317)

Wow, okay, so you’re saying that your limitations were that you don’t think you got enough aces with ASD?  No mention of the fact that your control group didn’t control anything, that you’re basing your entire interpretation of your results off of a theory your results disprove, that your ASD group was recruited from weirdly specific places and thus was probably already biased, or that your survey questions appear to believe that everyone goes immediately from hugging and kissing to having PIV?  I would say all of those are limitations.  Also, seriously, have you had any contact with the ace community, Gimour et al.?  Aces sign up for these sorts of surveys ALL THE TIME, and I’m sure there are more than a few aces with ASD who would have been happy to take your survey if it hadn’t been advertised in such weird places.

In conclusion:

Basically, we cannot conclude anything from this study about the connection between ASD and asexuality, because their survey design is kind of terrible.

But let’s assume for a moment that their results are correct, and people with ASD are more likely to be asexual than the general population.  This still does not mean that ASD causes asexuality any more than it means that ASD causes homosexuality (which they also found higher than average rates of) or that asexuality causes left-handedness, shortness, or late onset of menarche (other interesting correlations that researchers have found).

So what can we say about this study?  Well, maybe there are increased levels of asexuality in people with ASD.  Maybe there aren’t.  If there were, it’s more likely that ASD and non-heterosexuality have a cause in common than that ASD causes non-heterosexuality.

Also, seriously, Gilmour et al., take a class on survey design.

*Special thanks go to codeman38 and metapianycist for their suggestions on ASD-related readings, codeman38 and my partner for reading earlier drafts of this and suggesting improvements, and Sciatrix for joining me in a session of survey design critiquing that spawned much of the commentary on this piece.

*More readings on Extreme Male Brain theory and why it kind of fails:

Delusions of Gender by Cordelia Fine

Nature ran a piece on scientists and autism, which discusses Extreme Male Brain theory among other things.

And here’s a blog by an autistic woman about why she’s skeptical about the theory.

About queenieofaces

QueenieOfAces is a graduate student in the U.S. studying Japanese religion. She is a queer asexual. She also blogs over at Concept Awesome and runs Resources for Ace Survivors. She is never quite sure what to write in these introduction things, but this one time she accidentally got a short story on asexuality published in an erotica magazine.
This entry was posted in Research. Bookmark the permalink.

8 Responses to Paper summary: Sexuality in a community based sample of adults with autism spectrum disorder

  1. Jasleh says:

    wow…
    Anecdotal evidence here, but I have Aspergers (ASD) (also, I’m ace). Before realizing I was ace I dated two other people with Aspergers – both very much not ace (one bi, one straight). Soo… tiny sample size, but I’m not convinced there’s a higher number of aces in the ASD population. I do think there’s a higher number of ASD people in online forums, which may make it look that way.
    Also, this sounds like a really terrible study -_-

  2. Ember Nickel says:

    Huh, pity about the methodology, I’d have loved to read more of this.

  3. Seth says:

    Accusing them of implying that PiV is the only way to go further comes across as a little backward when by asking those two questions separately, they’re acknowledging that there are other ways, but PiV is the only one they specifically care about. Still bad, but for different reasons.

    This does seem like a laughably bad paper, overall. And I’m feeling a bit insulted by that EMB theory.

    • Siggy says:

      But Queenie said that they asked these two different questions, and then measured the correlation as a way of determining if they understand sexual language. According to the sexual researchers, if you give different answers to the different questions, that just means you didn’t understand them correctly. At least, that is what I am reading from Queenie’s summary.

      • queenieofaces says:

        Siggy is correct. The issue isn’t that they asked about PiV–the issue is that those who had not had PiV but HAD gone further than necking were marked as not understanding sexual language. If you’re attempting to measure a population that you assume has a larger than average proportion of homosexual and bisexual people, using PiV experience as a measure of knowledge of sexual language will backfire pretty badly!

  4. Siggy says:

    Aces sign up for these sorts of surveys ALL THE TIME, and I’m sure there are more than a few aces with ASD who would have been happy to take your survey if it hadn’t been advertised in such weird places.

    I don’t think they’re saying that they had too few ace respondents and needed more. Rather, they’re saying that asexuals may have been less likely to respond, and therefore they may be underestimating the number of asexuals in the target group.

    I think they got it backwards though. Asexuals in general may be less likely to respond to sexuality surveys, but I bet the opposite is true of the subset of asexuals who self-identify as such. After all, asexuals are encouraged to be vocal and represent. I think asexual awareness is relatively high in online autistic communities, and therefore a relatively large number of autistic asexuals may be self-identifying as asexual.

    • queenieofaces says:

      This is true; I didn’t phrase that all too well. I meant that self-identified aces were VERY likely to respond to these sorts of surveys (and apparently part of the survey had to do with self-identification of orientation, although they don’t show those survey questions, unfortunately). It’s still kind of strange that they didn’t advertise in any of the autistic communities/sites with large asexual contingents (and apparently don’t know about self-identified ace guerilla survey-taking culture).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s